1 / 36

NASA OBPG Radiometric Corrections for MODIS Aqua Ocean Color Bands

NASA OBPG Radiometric Corrections for MODIS Aqua Ocean Color Bands. Gerhard Meister NASA Code 616 Goddard Space Flight Center, MD 12/05/11 NOAA Science Center, World Weather Building, Rm 707, Camp Springs, MD Ocean Color Science Team Meeting. 1. 1. Ocean Biology Processing Group:.

luigi
Download Presentation

NASA OBPG Radiometric Corrections for MODIS Aqua Ocean Color Bands

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. NASA OBPG Radiometric Corrections for MODIS Aqua Ocean Color Bands Gerhard Meister NASA Code 616 Goddard Space Flight Center, MD 12/05/11 NOAA Science Center, World Weather Building, Rm 707, Camp Springs, MD Ocean Color Science Team Meeting 1 1

  2. Ocean Biology Processing Group: • NASA Code 614.2, Ocean Sciences Research, Hydrospheric and Biospheric Sciences Research, Goddard Space Flight Center • Responsible for producing Ocean Color (OC) products at NASA (CZCS, SeaWiFS, MODIS Aqua and Terra, MERIS, etc.) • Website: oceancolor.gsfc.nasa.gov

  3. MODIS design: • Rotating scan mirror • 10 (or more) detectors per scan line • Calibration and polarization characterization is scan angle, mirror side, and detector dependent

  4. MODIS issues: • Small calibration errors (~0.1%) lead to striping/erroneous trending in OC products • MODIS scan angle radiometric sensitivity changes with time (not for SeaWiFS) • On-orbit calibration can only be trended for lunar view angle (beginning of scan) and solar diffuser view angle (2nd half of scan, see later slide) • No on-board capability to trend polarization sensitivity changes on-orbit (not an issue yet for MODIS Aqua, but for MODIS Terra) 4

  5. MODIS Optical System Solar Diffuser view Lunar view

  6. Sensor degradation: SeaWiFS

  7. Sensor degradation: MODIS band 8 (MCST) 412nm

  8. Sensor degradation: MODIS band 9 (MCST) 443nm

  9. Sensor degradation: MODIS band 12 (MCST) 547nm

  10. Temporal issues: 412nm and red bands 10 10

  11. RVS issues: 412nm, red bands 11 11

  12. Approach summarized: • Baseline: MCST lunar and SD trending (coll. 6) • New MCST lunar analysis: time dependent NIR RVS • MODIS Aqua crosscalibrated to SeaWiFS (as for Terra in Kwiatkowska et al., Applied Optics, 2008) • Approach: Use SeaWiFS L3 nLw, bring to TOA, adjust MODIS calibration for every month of the mission (4-day L3) • Verify with analysis using only MODIS Aqua data: temporal trends (seasonal cycle removed) and ratio of L2/L3 versus scan angle 12 12

  13. MCST coll. 6 resolves temporal issue for 678nm, FLH: 13

  14. MCST coll. 6 reduces temporal variation for 488-547nm: 14

  15. MCST coll. 6 increases trend in 412nm, improves angstrom: 15

  16. Lt Qt Ut 0 Modeling of TOA Stokes vector over oceans air whitecap glint gas aerosol water Lt(λ) = [ Lr(λ) + La(λ) + tLf(λ) + TLg(λ) + td(λ)Lw(λ) ] · tg(λ) from MODIS NIR assumes MCST NIR band characterization λ’ → λ fit based on bio-optical models SeaWiFS 4-day mean Lwn(λ’)

  17. Crosscalibration approach: Lm = M11*Lt + m12*Q + m13*U Lm: measured TOA radiance (MODIS) Lt: true TOA radiance (from SeaWiFS) Q, U : linear Stokes vector components, modeled from Rayleigh and glint M11, m12, m13 : fitted instrument characterization parameters (depend on band, MS, detector, scan angle) 17

  18. Crosscalibration results: Polarization (temporal) - Larger seasonal cycle than in MODIS Terra - No trend in polarization coefficient m12 until 2008, not clear if trend afterwards Band 8: 412nm Black, blue, red:detectors 1,5,10, MS 1 View angle: nadir 18 18

  19. Crosscalibration results: Polarization (temporal) • - Cycle in m12 decreases with wavelength • All bands stable over time • Detector variation agrees with prelaunch data Band 10: 488nm Band 13: 667nm 19 19

  20. Crosscalibration results: Polarization (temporal) - Variability in m13 similar as in Terra - All bands stable over time, prelaunch values used Band 8: 412nm Band 13: 667nm 20 20

  21. Crosscalibration results: Calibration (m1 and RVS) • Implementation for 2010 reprocessing: Temporal correction for 412-443nm, constant correction for 488-678nm 21 21

  22. Comparison to MODIS Terra polarization sensitivity as a function of time at different view angles: • Significant variations in the blue (up to 40% at the end of scan at 412nm (band 8)), very small corrections for the red (band 13 at 667nm) Color coding: Frames (pixels) 22, 675, 989, 1250(out of 1354) Solid line is a fit to the measurements of each month (diamonds)

  23. Comparison to MODIS Terra radiometric corrections as a function of time at different view angles: • Significant corrections in the blue (up to 10% at 412nm (band 8)), very small corrections for the red (band 13 at 667nm) Color coding: Frames (pixels) 22, 675, 989, 1250 (out of 1354) Solid line is a fit to the measurements of each month (diamonds)

  24. RVS issues resolved: 412nm, 443nm 24 24

  25. RVS issues resolved: red bands (not EOM) 25 25

  26. Temporal issue resolved: 412nm 26

  27. Chlorophyll trends: similar to SeaWiFS 27

  28. Summary • Principal changes for MODIS Aqua calibration and characterization:- New temporal NIR scan angle dependence (MCST) - New temporal trends for 412-443nm (xcal)- New scan angle dependence for 488nm-678nm (xcal, no time dependence)- New detector dependence of polarization sensitivity (prelaunch) • Resulting improvements to ocean color products:- FLH stable over mission in olig.- Rrs 412nm stable over mission, variability reduced for remaining bands- Large scan dependence at EOM removed for 412nm- Minor scan angle dependence removed for 443-547nm- Large scan angle dependence reduced for 667-678nm, but still present at EOM 28

  29. Backup 29

  30. Lm Qm Um 0 M11 M12 M13 M14 M21 … M24 M31 … M34 M41 M42 M43 M44 Lt Qt Ut 0 = · Vicarious TOA MODIS total signal Lm(λ) = M11Lt(λ) + M12Qt(λ) + M13Ut(λ)

  31. TOA sensor cross-calibration MODIS measured TOA radiance, polarized m Lm(λ) – M11Lt(λ) + M12Qt(λ) + M13Ut(λ) • minimize over global distribution of path geometries • find best M11, M12, M13 per band, detector, and mirror-side • M11, M12, M13 = f (mirror AOI) • do this for one day per month over the mission lifespan

  32. Polarization of the atmosphere √Qt2 + Ut2 dp = Lt degree of atmospheric polarization dp air molecule (Rayleigh) and glint scattering polarization correction fp pre-launch MODIS characterization MODIS Terra swath 412nm band 8 Lm fp = Lt

  33. Issues and limitations with MODIS-Terra • Temporal and scan-dependent trend in Lwn of MODIS-Terra suggests • instrument RVS may be in error • polarization sensitivity may be changing • on-board calibration capabilities may be degrading (SD, SDSM) • On-board calibration (lunar or solar) CANNOT assess • changes in RVS “shape” • changes in polarization sensitivities • OBPG developed a vicarious approach for on-orbit characterization of Terra • RVS, and • polarization sensitivity

  34. New lunar trending of bands 13-16 (667-869nm) Comparison of collection 5 LUT to coll. 6: 34

  35. Crosscalibration results: Polarization (detectors) • Crosscalibration results confirm detector trend from prelaunch measurements (not used before) - Absolute offset at BOS (low TOA deg. of pol.) Band 8: 412nm Band 12: 547nm 35 35

  36. Surface Effects Sun Glint Corrections based on statistical models (wind & geometry) White Caps

More Related